Impact of changes in mortality on FRAX-derived fracture probabilities

Abstract

BackgroundAccurate hip fracture incidence and mortality rates are two essential requirements for FRAX calculators. PurposeTo investigate the effects of change in mortality on FRAX-derived fracture estimates. MethodsLebanese FRAX calculator was updated in 2012 from version 3.00 utilizing WHO mortality data from year 1999, and hip fracture incidence rates from 2007, to version 3.05 utilizing mortality data from 2009, but with identical hip fracture data. FRAX-derived estimates from 679 patients were computed using both FRAX versions and compared. Numbers presented as median [25th–75th] percentiles. ResultsThe 10-year FRAX-derived probability of major osteoporotic fracture and hip fracture increased substantially. Changes were most pronounced in high risk sub-groups. The relative increase in probability of major osteoporotic fracture in individuals with a baseline risk of 10–20% was 79% [19%–127%], and in individuals with a baseline risk >20% it was 125% (N=3). The numbers for relative increase in hip fracture probability were 98% [33%–135%], and 129%, respectively. Similarly, individuals older than 70years had a 125% [89%–150%] relative increase in probability of major osteoporotic fracture and a 122% [95%–145%] relative increase in hip fracture probability. Using the FRAX-based Lebanese guidelines, FRAX 3.05 led to an additional increase in treatment qualification of 3.8 patients per 100 patients, or a relative increase of 24%. ConclusionsUpdates in mortality values increased FRAX-derived estimates, most substantially in older patients, and those at high risk for fracture. The update results in altering individuals' treatment decisions and modifying country wide osteoporosis management. Our results are relevant to the development and update of FRAX models for countries worldwide, and more importantly those with increasing longevity and possible increase in fracture rates.